2014 UT Schedule Analysis....

[fur burglar]

Science mother f***er.

:stop: All that jibber jabber wasn't scientific, it was statistical, dumb-butt

 

[WilliamDuVOL]

:stop: All that jibber jabber wasn't scientific, it was statistical, dumb-butt

I very much appreciated the post for its statistical value, I just felt my statement punctuated it perfectly.

 

[sjt18]

The over/under win total predictions came out today, I think from a Vegas sports book, not sure the source.....they put UTs total at 4.5. For comparison, Florida's # was 7, Vandy's was 6.

Do you gamble? If I did, I'd put money on "over"... and might even put money on Vandy for under.

 

[sjt18]

I find it hard to believe we can win USC, Ole Miss, and Mizzou...

You have set your expectations so low that you would find it hard to believe that UT could beat Carson-Newman's club team.

 

[KBVol]

Do you gamble? If I did, I'd put money on "over"... and might even put money on Vandy for under.

I don't, but I agree 100%.

 

[sjt18]

We pull off a close on against Utah State. This game reminds me of the Air Force game in which we won on the last defensive play of the game. the q.b. ran for a two point play and Mitchell shed the block and tackled the q.b. on the 3 yard line to secure the win. That win was huge because it got us in a fairly good bowl game against Wisconsin.

We will easily beat Arkansas State.


We will easily beat Chattanooga.


We will beat vandy.

We will beat an improved kentucky.

We lose to UGA, USC, Misery, Ole Piss, " O Bama" Florida & Oklahoma

Miss a bowl game again but get the freshmen a year of valuable experience.....

So you are excusing a season in which UT loses to 3 opponents with less talent, correct? In seriousness, I don't quite buy daj's info as ironclad but UT SHOULD beat Mizzou. They simply lost too many of the guys who made them competitive and cannot replace them with comparable talent.

 

[TCbodybuilder]

People are hyperventilating over Utah State and Arkansas State, why?

Because bowl games, and the records to get there, are touted as meaningful, when in fact they are meaningless. The statistical likelihood is that UT could have done just as well against Utah State's schedule, if not better, than the Aggies did. Conversely, Utah State would have likely done far worse than 5-7 with UT's schedule.

If talent averages predict about 70% of all games played (and they do, actually that number is higher when one team recruits in the top 50 and the other recruits in the bottom 50), then one should only look at Utah State's 2013 schedule versus Tennessee 2013 schedule to see why UT went 5-7, and Utah State won 9.

View attachment 77634

Here is my explanation:

Using recruiting evaluations as a guide, the mighty Utah State Aggies should have only won 4 games, but they won 9 including a bowl game against Northern Illinois (91). It is a key point that teams that play each other that are both in the bottom 50 tend to have a less stable correlation between talent and wins. this is likely due to the fact that recruiting services are really good at distinguishing between 3 to 5 star players and really mediocre at distinguishing unranked through high two star players (the type of players that Utah State recruits). Similarly, the probability of victory is significantly higher than 70% when a team who recruits in the top 50 plays a team in the bottom 50. Back to last year's Utah State schedule: the best team that they beat, viewed through a talent lens, was no better than 80th (Hawaii). Their worst loss also came against a team ranked about 80th (Fresno State). Importantly, Utah State's average Division 1 caliber win came against an opponent who was roughly ranked 93rd. Their average loss came against a team ranked 48.6.

For comparison, Tennessee should have won 7 games, but won 5. The Vol's best win came against a South Carolina team ranked about 18th using recruiting evaluations, and the worst loss came against a Vandy team ranked about 48th. UT's average Division 1 caliber win was against a team ranked 51, and Tennessee's average loss came against a team ranked 16.6.

In other words, Utah State's best win came against a team ranked 32 rungs below Tennessee's worst loss. Imagine trying to climb a ladder that is missing 32 rungs between what you are capable of on your best day, to what your opponent is capable of on their worst. The challenge of making that leap isn't impossible, but it is highly unlikely.

Or for a little different sort of comparison, Tennessee average loss came against teams that are built like Oregon, Texas A&M, or UCLA (those teams fall around the 16 ranking in class averages) whereas Utah State averaged losing to teams built like Kentucky, Kansas, and North Carolina State.

Admittedly that is last year. Tennessee's roster in 2014 has an explosion of young talent, and Utah State's roster actually fell from 101 to 103. Yes, the Vol's will field a young team in the season opener (see below for change in talent by position group) but Utah State is filling some holes on their roster as well.

View attachment 77635

EDIT: If you want a look at how Tennessee's talent ranks in the 2014 schedule, here is a chart to illustrate that point.

View attachment 77638

All games indicated in red are a 70% chance for a loss, and green are a 70% chance of a win. My data shows that as a general rule roughly 80% of teams in the SEC will fall within 2 games of these predicted win/loss matrices.

What is your full data set, analysis methods, and software you are using?

 

[KnoxRealtorVOL]

OK is overacted...Like every year

Yes, and Alabama will be a much less talented team than we've seen in recent years.

........ both teams are still going to destroy us. Got to win the swing games this year. Need another 2013 SC-type win and got to beat Vandy/KY.

As much as none of us want to admit it, not winning both of the Vandy/KY games has been the reason we haven't gone bowling since 2010.

 

[sjt18]

Me too.

And statistically there is only a 34% chance we win all three, similarly there is only a 2% chance that we lose all three.

How do you figure those probabilities? That doesn't seem to be the correct methodology.

 

[sjt18]

I agree. The caveat however is that while Vandy, once again, has the SEC's easiest schedule, UT, once again, has the conference's most difficult.

Easiest? I would have thought that would be Mizzou.

UF's I believe was rated the most difficult followed by UT. UF gets LSU, Bama, and FSU this year plus their East games.

 

[KBVol]

So you are excusing a season in which UT loses to 3 opponents with less talent, correct? In seriousness, I don't quite buy daj's info as ironclad but UT SHOULD beat Mizzou. They simply lost too many of the guys who made them competitive and cannot replace them with comparable talent.

Oh yeah, over/under number for Mizzou was 9... or double UTs. Why do I think you're calling your bookie right now?

 

[KBVol]

Easiest? I would have thought that would be Mizzou.

UF's I believe was rated the most difficult followed by UT. UF gets LSU, Bama, and FSU this year plus their East games.

Not according to the guys I heard discussing this this morning. Said once again that Vandy had the easiest and UT the toughest.

 

[TrueOrange]

Easiest? I would have thought that would be Mizzou.

UF's I believe was rated the most difficult followed by UT. UF gets LSU, Bama, and FSU this year plus their East games.

Mizzou gets Arky and @A&M

Vandy gets the two Mississippi schools.

I'd have to go with Vandy there.

 

[sjt18]

OK. We're probably splitting hairs. Vandy is near or at the bottom and UT is near or at the top in SOS.

 

[KBVol]

OK. We're probably splitting hairs. Vandy is near or at the bottom and UT is near or at the top in SOS.

Yes, and keep in mind the OOC games. That's where Vandy falls completely off the cliff.

 

[daj2576]

What is your full data set, analysis methods, and software you are using?

Rivals recruiting numbers back to 2002 meaning first 4 year average begins in 2005. Other than that it's lots of time in excel. If you want fully regression tested figures, check Dave Bartoo's website (cfbmatrix.com). I developed this before I knew he existed but his methodology is similar, older, and tested. Our results and opinions are close, though we depart on certain details. Hes a great guy and fun to talk to.

 

[daj2576]

Science mother f***er.

I read that as "SILENCE mother....". It still made perfect sense in that context as that is the typical reaction I get. Lol

 

[daj2576]

:stop: All that jibber jabber wasn't scientific, it was statistical, dumb-butt

What exactly do you find to be "statistical, dumb-butt"?

 

[daj2576]

So you are excusing a season in which UT loses to 3 opponents with less talent, correct? In seriousness, I don't quite buy daj's info as ironclad but UT SHOULD beat Mizzou. They simply lost too many of the guys who made them competitive and cannot replace them with comparable talent.

I've given probabilities which are not "iron clad" by definition.

I totally get why most buck what I'm saying. Most tend to over represent exceptions to a rule (App State v Michigan will typically be cited as an example of how this doesn't work) as opposed to the rule. Also, most UT fans have experienced being the exception for the better part of a decade, and tend to extrapolate outwards from that (inductive vs deductive reasoning). Both views, when compared to mountains of games played, are incorrect. I can't force anyone to believe, nor do I want to. The debate is part of the fun.

On the other hand I am a bit disheartened by how many people (present company excluded) are terrified to challenge what they know. We have reached an age where information is sought to affirm our position instead of inform us to make an opinion. C'est la vie.

 

[GBO97]

We pull off a close on against Utah State. This game reminds me of the Air Force game in which we won on the last defensive play of the game. the q.b. ran for a two point play and Mitchell shed the block and tackled the q.b. on the 3 yard line to secure the win. That win was huge because it got us in a fairly good bowl game against Wisconsin.

We will easily beat Arkansas State.


We will easily beat Chattanooga.


We will beat vandy.

We will beat an improved kentucky.

We lose to UGA, USC, Misery, Ole Piss, " O Bama" Florida & Oklahoma

Miss a bowl game again but get the freshmen a year of valuable experience.....

If we can beat SC last year, no way in heck we can't beat at least 2 of the 7 teams you listed as L's (many of which will be down from last year).

Also the AF win was 2006, the season when we played vs Wisconsin in the bowl was 07.

 

[Catbone]

What exactly do you find to be "statistical, dumb-butt"?

Daj what are your thoughts on us winning 8 games based on talent and Vegas at 4.5? I found that interesting

 

[I Love Lamp]

People are hyperventilating over Utah State and Arkansas State, why?

Because bowl games, and the records to get there, are touted as meaningful, when in fact they are meaningless. The statistical likelihood is that UT could have done just as well against Utah State's schedule, if not better, than the Aggies did. Conversely, Utah State would have likely done far worse than 5-7 with UT's schedule.

If talent averages predict about 70% of all games played (and they do, actually that number is higher when one team recruits in the top 50 and the other recruits in the bottom 50), then one should only look at Utah State's 2013 schedule versus Tennessee 2013 schedule to see why UT went 5-7, and Utah State won 9.

View attachment 77634

Here is my explanation:

Using recruiting evaluations as a guide, the mighty Utah State Aggies should have only won 4 games, but they won 9 including a bowl game against Northern Illinois (91). It is a key point that teams that play each other that are both in the bottom 50 tend to have a less stable correlation between talent and wins. this is likely due to the fact that recruiting services are really good at distinguishing between 3 to 5 star players and really mediocre at distinguishing unranked through high two star players (the type of players that Utah State recruits). Similarly, the probability of victory is significantly higher than 70% when a team who recruits in the top 50 plays a team in the bottom 50. Back to last year's Utah State schedule: the best team that they beat, viewed through a talent lens, was no better than 80th (Hawaii). Their worst loss also came against a team ranked about 80th (Fresno State). Importantly, Utah State's average Division 1 caliber win came against an opponent who was roughly ranked 93rd. Their average loss came against a team ranked 48.6.

For comparison, Tennessee should have won 7 games, but won 5. The Vol's best win came against a South Carolina team ranked about 18th using recruiting evaluations, and the worst loss came against a Vandy team ranked about 48th. UT's average Division 1 caliber win was against a team ranked 51, and Tennessee's average loss came against a team ranked 16.6.

In other words, Utah State's best win came against a team ranked 32 rungs below Tennessee's worst loss. Imagine trying to climb a ladder that is missing 32 rungs between what you are capable of on your best day, to what your opponent is capable of on their worst. The challenge of making that leap isn't impossible, but it is highly unlikely.

Or for a little different sort of comparison, Tennessee average loss came against teams that are built like Oregon, Texas A&M, or UCLA (those teams fall around the 16 ranking in class averages) whereas Utah State averaged losing to teams built like Kentucky, Kansas, and North Carolina State.

Admittedly that is last year. Tennessee's roster in 2014 has an explosion of young talent, and Utah State's roster actually fell from 101 to 103. Yes, the Vol's will field a young team in the season opener (see below for change in talent by position group) but Utah State is filling some holes on their roster as well.

View attachment 77635

EDIT: If you want a look at how Tennessee's talent ranks in the 2014 schedule, here is a chart to illustrate that point.

View attachment 77638

All games indicated in red are a 70% chance for a loss, and green are a 70% chance of a win. My data shows that as a general rule roughly 80% of teams in the SEC will fall within 2 games of these predicted win/loss matrices.

If we go 8-4 there will be a mattress shortage in the state of Tennessee.

Just curious....do you happen to have Auburn's predicted record from last season using your method?

 

[Neyland Law Vol]

 

[daj2576]

If we go 8-4 there will be a mattress shortage in the state of Tennessee.

Just curious....do you happen to have Auburn's predicted record from last season using your method?

They performed exactly as talent predicted (the loss to lessor LSU and win against a better Bama team was a wash) even down to losing to FSU in national champ game.

I was screaming from the rafters to beware of Auburn, but that has nothing to do with Malzahn and everything to do with talent.

There is a link in this link that might interest you. http://www.volnation.com/forum/tennessee-vols-football/218378-finebaum-show-2.html#post10179056

 

[daj2576]

Daj what are your thoughts on us winning 8 games based on talent and Vegas at 4.5? I found that interesting

Vegas is good at what they do especially when viewed from a birds eye view. My limited study shows that as far as pure winners and losers (ignore the spread and stick with who they pick to win), they are right about 80% of the time. However, when viewing an over/under that isnt the same sort of calculus. It seems to be a line to entice betting, much like a spread. I've never studied how correct the seasonal predictions are in Vegas and I bet that line moves due to betting like a spread. It is possible that they are correct and that UT will once again be an outlier where talent doesn't predict success. Still, these recruiting eval numbers and insofar as seasonal predictions, are right north of 80% of the SEC (likely this holds true for all big 5 conferences) within 2 games a season. Individually, each game is 70/30 in favor of the team with a higher recruiting average.